Fu-Jen Chu
Papers
15
Total Citations
742
H-Index
8
About
Fu-Jen Chu is a robotics and computer vision researcher whose work sits at the intersection of deep learning, robotic manipulation, and assistive technology. He is best known for pioneering contributions to grasp detection and affordance understanding, two foundational challenges in enabling robots to interact meaningfully with the physical world. Chu's most influential work, "Real-World Multiobject, Multigrasp Detection" (2018), has garnered over 447 citations and introduced a novel deep learning architecture that reframes grasp prediction as a classification problem using null hypothesis competition, significantly improving robustness in cluttered, real-world scenes. Building on this foundation, he developed frameworks for affordance segmentation, keypoint detection, and primitive shape decomposition, enabling robots to reason not just about where to grasp, but how object parts can be functionally used — often leveraging synthetic training data to bypass costly real-world annotation. Beyond perception, Chu has extended his research into assistive robotics, designing hands-free manipulation systems that combine augmented reality with tongue-drive interfaces to empower individuals with physical disabilities. With a cumulative citation count exceeding 700, his body of work reflects both technical depth and a clear commitment to translating robotic perception research into systems that meaningfully improve human lives.
Research Focus
Key Achievements
Top Papers
- 1Real-World Multiobject, Multigrasp Detection447 citations · 2018
- 2
- 3An Affordance Keypoint Detection Network for Robot Manipulation52 citations · 2021
- 4
- 5Deep Grasp: Detection and Localization of Grasps with Deep Neural Networks.34 citations · 2018
- 6
- 7
- 8Real-world Multi-object, Multi-grasp Detection14 citations · 2018
- 9
- 10